Tree Seed Optimization Algorithm Based On Levy Flight And Its Application In Image Segmentation

The Tree Seed Optimizer Algorithm(TSA)is an emerging group intelligent optimization algorithm that imitates the seeds of trees in nature and grows up.It has been proven that TSA can provide relatively competitive accuracy results.Because of its flexibility,efficiency and robustness,the tree seed optimization algorithm not only attracts many researchers to improve it,but also these researchers uses the tree seed intelligent optimization algorithm in many practical applications.However,with the deepening of the research,some scholars have found that the algorithm has the disadvantages of poor precision and slow convergence.In this paper,some improvements are proposed to make the process of solving the problem more effective.The improved TSA is firstly used to detect the performance of function optimization problems,and then to solve various engineering testing problems.Finally,it is used to find the initial clustering center in the fuzzy clustering image segmentation,The purpose is to further improve the theory of the tree seed optimization algorithm and expand its application range.The work of this paper is mainly divided into the following aspects:(1)The tree seed optimization algorithm based on Levy flight is proposed.The levy flight strategy is introduced into the tree species optimization algorithm,which greatly enhanced the diversity of the population and avoids falling in the local convergence in the later stage of the algorithm,and the proposed algorithm is successfully applied to the function optimization problem.(2)The tree seed optimization algorithm based on Levy flight is applied to solve various engineering testing problems,to validate the tree seed optimization algorithm based on Levy flight is superior.(3)The tree seed optimization algorithm based on Levy flight is applied to the image segmentation problem based on fuzzy clustering.Traditional fuzzy clustering image segmentation algorithm is difficult to determine the initial clustering center.Therefore,it is difficult to achieve efficient image segmentation.The tree seed algorithm based on Levy flight is applied to the image segmentation of fuzzy clustering,which makes the initial clustering center easier to determine and search,and helps to complete image segmentation.